From: David Hildenbrand <david@redhat.com>
To: "Wang, Wei W" <wei.w.wang@intel.com>, Peter Xu <peterx@redhat.com>
Cc: Hailiang Zhang <zhang.zhanghailiang@huawei.com>,
Juan Quintela <quintela@redhat.com>,
"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>,
"Dr . David Alan Gilbert" <dgilbert@redhat.com>,
Leonardo Bras Soares Passos <lsoaresp@redhat.com>
Subject: Re: [PATCH] migration: Move bitmap_mutex out of migration_bitmap_clear_dirty()
Date: Mon, 5 Jul 2021 15:41:52 +0200 [thread overview]
Message-ID: <a5877d58-d501-0ff6-676b-c98df44d1b6f@redhat.com> (raw)
In-Reply-To: <562b42cbd5674853af21be3297fbaada@intel.com>
On 03.07.21 04:53, Wang, Wei W wrote:
> On Friday, July 2, 2021 3:07 PM, David Hildenbrand wrote:
>> On 02.07.21 04:48, Wang, Wei W wrote:
>>> On Thursday, July 1, 2021 10:22 PM, David Hildenbrand wrote:
>>>> On 01.07.21 14:51, Peter Xu wrote:
>>
>> I think that clearly shows the issue.
>>
>> My theory I did not verify yet: Assume we have 1GB chunks in the clear bmap.
>> Assume the VM reports all pages within a 1GB chunk as free (easy with a fresh
>> VM). While processing hints, we will clear the bits from the dirty bmap in the
>> RAMBlock. As we will never migrate any page of that 1GB chunk, we will not
>> actually clear the dirty bitmap of the memory region. When re-syncing, we will
>> set all bits bits in the dirty bmap again from the dirty bitmap in the memory
>> region. Thus, many of our hints end up being mostly ignored. The smaller the
>> clear bmap chunk, the more extreme the issue.
>
> OK, that looks possible. We need to clear the related bits from the memory region
> bitmap before skipping the free pages. Could you try with below patch:
I did a quick test (with the memhog example) and it seems like it mostly
works. However, we're now doing the bitmap clearing from another, racing
with the migration thread. Especially:
1. Racing with clear_bmap_set() via cpu_physical_memory_sync_dirty_bitmap()
2. Racing with migration_bitmap_clear_dirty()
So that might need some thought, if I'm not wrong.
The simplest approach would be removing/freeing the clear_bmap via
PRECOPY_NOTIFY_SETUP(), similar to
precopy_enable_free_page_optimization() we had before. Of course, this
will skip the clear_bmap optimization.
--
Thanks,
David / dhildenb
next prev parent reply other threads:[~2021-07-05 13:45 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-06-30 20:08 [PATCH] migration: Move bitmap_mutex out of migration_bitmap_clear_dirty() Peter Xu
2021-07-01 4:42 ` Wang, Wei W
2021-07-01 12:51 ` Peter Xu
2021-07-01 14:21 ` David Hildenbrand
2021-07-02 2:48 ` Wang, Wei W
2021-07-02 7:06 ` David Hildenbrand
2021-07-03 2:53 ` Wang, Wei W
2021-07-05 13:41 ` David Hildenbrand [this message]
2021-07-06 9:41 ` Wang, Wei W
2021-07-06 10:05 ` David Hildenbrand
2021-07-06 17:39 ` Peter Xu
2021-07-07 12:45 ` Wang, Wei W
2021-07-07 16:45 ` Peter Xu
2021-07-07 23:25 ` Wang, Wei W
2021-07-08 0:21 ` Peter Xu
2021-07-06 17:47 ` Peter Xu
2021-07-07 8:34 ` Wang, Wei W
2021-07-07 16:54 ` Peter Xu
2021-07-08 2:55 ` Wang, Wei W
2021-07-08 18:10 ` Peter Xu
2021-07-02 2:29 ` Wang, Wei W
2021-07-06 17:59 ` Peter Xu
2021-07-07 8:33 ` Wang, Wei W
2021-07-07 16:44 ` Peter Xu
2021-07-08 2:49 ` Wang, Wei W
2021-07-08 18:30 ` Peter Xu
2021-07-09 8:58 ` Wang, Wei W
2021-07-09 14:48 ` Peter Xu
2021-07-13 8:20 ` Wang, Wei W
2021-07-03 16:31 ` Lukas Straub
2021-07-04 14:14 ` Lukas Straub
2021-07-06 18:37 ` Peter Xu
2021-07-13 8:40 ` Wang, Wei W
2021-07-13 10:22 ` David Hildenbrand
2021-07-14 5:03 ` Wang, Wei W
2021-07-13 15:59 ` Peter Xu
2021-07-14 5:04 ` Wang, Wei W
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=a5877d58-d501-0ff6-676b-c98df44d1b6f@redhat.com \
--to=david@redhat.com \
--cc=dgilbert@redhat.com \
--cc=lsoaresp@redhat.com \
--cc=peterx@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=quintela@redhat.com \
--cc=wei.w.wang@intel.com \
--cc=zhang.zhanghailiang@huawei.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).